Co2 leak? Help please!

[nero82]

Hey everyone,

I've posted this question on another forum to a deafening silence. Maybe some of you could help me? My spare bottle is being filled at the moment, and I'm hoping the current one will last me...

I just got a new reg with solenoid, but it seems to be chugging through the co2 3 times as fast as the old reg. Of course the first thing I thought was LEAK, but for the life of me I can't find one.

The bottle pressure is dropping about 5psi a day... it is a one liter tank...I think that is about 600g. At about 3 bps that seems way too much. A tank would barely last me 20days that way. last time the tank went over a month. In fact I just checked it again then, It looks to have gone down about 40psi over the last five days

I marked the position of the needle on the gauge, then left it overnight with the solenoid closed, and the needle didnt budge. I assume that rules out leaks in all the parts up to the fine adjustment, except if the solenoid is leaking in the ON position. I dunked the whole bubble counter in water, and found no leaks there. Soapy water didn't find any leaks where the hose joins the reg. I coated as much of the solenoid as I could in soapy water and couldnt find anything there either.

Can anyone point me in the right direction? Is this usage plausible without a leak? Is there anything else I can do to test for a leak? My other identical tank with a cheap reg and no solenoid lasted me a whole month.

Cheers,
Nero

 

[nero82]

Update: been poking around a bit more...

Are cylinders meant to hold pressure through most of their usage? That is to say, does the bottle gauge drop in a linear fashion, or does it stay up until the tank is depleted, then decrease rapidly? If that is the case, then I think I totally overdosed on the co2 for the first week... I cranked it up really high for a while there (got a bit overzealous) and I'm just seeing the gauge start to drop rapidly now?

Hope someone can shed some light. the more I read the more confused I get!

 

[nero82]

ooookay now I just feel foolish. Just read this on a homebrew forum. Can anyone confirm this for me?

"The observed pressure on your high pressure gauge will not change until the tank has less than 10-20% CO2 remaining. Until all the liquid has evaporated inside the tank, the pressure will remain constant. When the liquid is gone, you’ll observe the pressure gauge decrease fairly rapidly upon use. "

 

[nipat]

ooookay now I just feel foolish. Just read this on a homebrew forum. Can anyone confirm this for me?

"The observed pressure on your high pressure gauge will not change until the tank has less than 10-20% CO2 remaining. Until all the liquid has evaporated inside the tank, the pressure will remain constant. When the liquid is gone, you’ll observe the pressure gauge decrease fairly rapidly upon use. "

Yes, that is typical behavior.

1 litre CO2 tank? I've never used that so can't comment.
My tank contains 2 Kg of CO2 and lasts about 4-5 months.

 

[nero82]

Right, so If mine is 600g, one month would be a good time frame to aim for. I'll see what the refilled tank does now I've set the bubble rate to 2bps I spose.

Thanks nipat

 

[Matt F.]

Testing the unit to see if it will hold pressure while the solenoid is closed will only test to see if the solenoid is closing/sealing properly. Provided the regulator is set to a certain working pressure, the regulator before the solenoid will keep a constant flow and working pressure rate. If you have a leak. The working pressure gauge might not move even with a leak. Losing that much co2 in such a short time suggest you have a moderate leak somewhere. COuld be from anywhere. The needle on the tank pressure gauge will not move from 700-800 psi till you run out of liquid co2. Once this happens, the tank could run out of gas in 1-2 weeks (maybe longer). In other words, time for a refill at that point. Something is leaking.
5psi from a tank per day is a big leak. My 10@# CO2 tank lasts >8 months at 6-8 bps.

What type of regulator do you have? Pictures?

 

[gsjmia]

There are some very technical answers on this forum, but from my knowledge, the Co2 in the tank is a liquid. So long as there is some liquid in the tank the pressure gauge won't drop. The gauge starts to drop when the liquid runs out and the final gas in the tank starts to run out--the guage is not linear through the life of the gas in the tank.

I have a 10lb tank and when it is at the end its normal to see the guage fall noticably each day.

I recently had a leak and it was the "permaseal" brass gasket with O rings on both sides.
I left it in and used a standard fibre washer on the tank side. No leaks for 2 weeks and then the O ring on the regulator side went bad. Discarded the permaseal and just used the cheap fibre washer and no leaks..

Do a bubble check on all your fittings--you will have to make a nest of bubbles in your hand and cup it around one side of the fitting, then the other side. Leaks and bubbles are kinda of hard to find--if the leak is a strong one, it will blow through the bubbles and you may not notice. If its a slow leak, it will take maybe 30 seconds to notice a growth in one of the bubbles.

 

[nero82]

Thanks everyone for the replies. I think I'm starting to understand a bit more. I've put the new bottle on today, I'll see how things go now that I know what I am looking at.

Nero

 

[hbosman]

A one liter bottle lasting a month seems about right to me. I use 20 OZ tanks which are a little smaller and that's about how long they lasted on my 46 gallon bowfront.

 

[Matt F.]

A one liter bottle lasting a month seems about right to me. I use 20 OZ tanks which are a little smaller and that's about how long they lasted on my 46 gallon bowfront.

Man, didn't even recall the size of his co2 tank. Makes sense, I guess. But 5psi per day still seems excessive, but I don't know for sure.

 

[Jim Miller]

Can I assume from your 1L bottle that you aren't in the USA? If in the US I've found that a thick leak detection fluid available at HomeDepot is much better than soapy water for detecting leaks. It is much thicker and persists on joints longer allowing for better leak detection. An alternative would be to make a mixture of glycerin, soap and a minimal amount of water. You want something thin enough to pour but not run off.

Good luck

Jim

 

[nero82]

Thanks for the replies,
I'm now suspecting it is not a leak. I've had the new bottle on for a couple of days and the pressure is steady at 850psi. Hasn't budged.
I was hitting the co2 really hard on the last bottle... I've since sat down and timed it, and it turns out one bubble per second is slow. reaaaaly slow. Who would have thought So I've cranked it back a bit to about 2 bubbles per second (accurately this time) And I'll see how long this bottle lasts.

Can I assume from your 1L bottle that you aren't in the USA?

Correct, I'm in Australia.

I did make up a solution from detergent, glycerin and a tiny bit of water. Found nothing. I sealed the threads with thread tape. I've replaced the check valve.

I'll let you know how it goes, thanks again for all the help!

Nero

 

[Jim Miller]

I like the thicker mix. I had a tiny leak where the CGA-320 stem entered the regulator body. I hadn't noticed it before when I did some initial leak testing that found more major ones. But when I did a long term test the high pressure gauge drifted down over several days. Armed with the thicker fluid I found that point produced a tiny bit of fizzy bubbles, not big ones and certainly not quickly.

Another few degrees with a wrench closed that one down permanently.

I've left the unit to see what other things happen over time as it isn't in operation. I've noticed that the low pressure gauge has now crept up to 40psi from it's initial reading while the high pressure gauge is drifting down. This implies a small leak of the high pressure side into the low pressure side. Nothing to worry about since it has taken several weeks to get this far.

Jim

 

[danielo]

Thanks for the replies, they've helped me greatly.